The clinical and biochemical profile of acquired C1 esterase inhibitor (C1-INH) deficiency, also known as acquired angioedema (AAE) syndrome, has been described in association with B cell lymphoproliferative disorders (1–4) and, less commonly, with autoimmune diseases, especially systemic lupus erythematosus– (SLE) like syndromes (1, 4–6). Two types of AAE have been proposed based on the immunopathologic origin of the deficiency. AAE-1 is usually associated with a lymphoma and may result from idiotype-antiidiotype antibody interactions leading to excessive complement activation and consumption of C1 and C1-INH (2). In AAE-2, autoantibodies block C1-INH function (3, 7). With increasing clinical recognition of AAE, and more sensitive tests for the detection of autoantibodies to the C1-INH, the distinction between these two conditions has blurred (3, 8, 9). For example, in some patients the paraprotein is an autoantibody that inhibits the function of the C1-INH (3).
We report 2 patients with C1-INH deficiency secondary to an autoantibody that presented diagnostic difficulties because of the lack of typical angioedema, and clinical syndromes and laboratory profiles suggesting SLE.
Case report 1
A 57-year-old white woman, with a history of Hashimoto thyroiditis, was referred for an analysis of the basis of C4 deficiency in the presence of antinuclear antibody– (ANA) negative lupus. In the early 1990s, she developed a clinical syndrome consisting of fatigue, episodic difficulty with cognition, intermittent crampy abdominal pain, and panic attacks. In November 1998, she presented with light-headedness, and a hemoglobin of 4.5 gm/dl secondary to autoimmune hemolytic anemia (AIHA). The direct antiglobulin test was positive for IgG and C3, and both cold and warm reactive autoerythrocyte antibodies were present. Tests for ANA, double-stranded DNA (dsDNA), anti-Ro/SSA and anti-La/SSB were negative. Antiphospholipid antibodies (aPL) were present. C4 was repeatedly undetectable, but C3 was normal. After pulse methylprednisolone therapy (500 mg twice daily for 3 days), she was discharged and prescribed daily oral prednisolone (1 mg/kg) and azathioprine (100 mg). The hemoglobin normalized after 8 weeks.
In March of 1999, when we first saw this patient, the AIHA was in remission. The patient reported increased distractibility, irritability, and worsening cognitive defects. She denied photosensitivity, alopecia, Raynaud's phenomenon, malar or discoid rash, renal disease, bleeding/clotting phenomena, or recurrent fetal loss. Although she reported no peripheral angioedema, over the past year she had experienced 5 episodes of crampy abdominal pain associated with malaise, nausea, and bloating. A gastroenterologist diagnosed intermittent pseudoobstruction. Family history was significant for a mother with Hashimoto thyroiditis, a brother with sarcoidosis, an aunt with Wegener's granulomatosis, and a son with recurrent idiopathic erythema nodosum. The complement profile (Table 1) was not consistent with isolated total C4 deficiency, but indicated classical pathway activation.
Results are expressed as the % of normal pooled human plasma unless otherwise noted. ND = not done.
Antigenic analysis of C5, C6, C7, C8, C9, Factor B, Factor I, Factor H and C4 binding protein, were unremarkable. Alternative pathway CH50 was normal. C1r and C1s were reduced, comparable to the C1q.
In many clinical immunology laboratories, a low C4 level is reported as less than 10 mg/dl. The level could be from 0 to the lower detection limit of the assay. The sensitivity of the assay can be adjusted to detect such levels. In patient 1, the levels were 0.4 to 1.5 mg/dl (0.4–2% of normal pooled plasma). The low but easily detectable CH50 in patient 1 confirms that functionally intact C4 was present.
This study was performed at the National Jewish Medical and Research Center, Denver, CO.
In November 1999, she was again treated successfully with high doses of corticosteroids for a relapse of the AIHA. Sequential courses of intravenous immunoglobulin, cyclosporin, and mycophenolate did not permit a reduction in the steroid dose. By January 2000, an IgM autoantibody to C1-INH was identified (Table 2). In March 2000, a lymphoplasmacytoid lymphoma (immunocytoma) was identified. Following a splenectomy (25 cm containing tumor), she developed a pulmonary embolism and worsening hemolysis requiring 14 transfusions. Treatment with 8 cycles of rituximab successfully induced a remission. During the past 2 years the C4 level has remained very low to undetectable, and the patient has not experienced peripheral angioedema or recurrent abdominal pain.
Table 2. Autoantibody profile
These studies were performed at the University of Milan, Italy.
Anti-C1-INH antibodies were measured by enzyme-linked immunosorbent assay, based on the method described by Alsenz et al (7). Results are expressed in arbitrary units (AU).
Anti-C1q antibodies were measured as above except that the microtiter plates were coated with purified C1q, and binding of plasma immunoglobulin was performed at high ionic strength (phosphate buffered saline containing 1 M NaCl).
A 55-year-old white woman was referred in November 1998 because of C1-INH deficiency. The defect had been identified 1 month earlier by a dermatologist during a consultation for eczema. A C1-INH test was performed because of 2 bouts of possible angioedema. In 1991, she had deep and superficial venous thrombosis, respectively, of the left leg, occurring 2 months apart. She was treated with oral anticoagulation for 6 months. From 1988 to 1998, her clinical history had been dominated by angina pectoris, but with normal coronary angiography. She was diagnosed with syndrome X (microvascular angina) and eventually improved after treatment with an angiotension converting enzyme inhibitor (gallopamil) and alpha-blocker (doxasosin).
The complement profile was typical for AAE (Table 1), and an autoantibody to the C1-INH was identified (Table 2). ANA and aPL antibodies were present. Tests for ENA, DNA, mitochondrial and smooth muscle antibodies were negative. During a two-year followup, she has had no further episodes of thrombosis or angioedema, and a lymphoproliferative disorder has not developed.
Angioedema was not a prominent aspect of either of these patients' clinical course and the diagnosis of AAE was established based on characteristic abnormalities in the complement profile. For patient 1, the complement measurements were part of an evaluation for presumed lupus. Evolution to a non-Hodgkin's lymphoma occurred, a common outcome in AAE. In patient 2, complement studies were part of a workup for possible angioedema. Positive tests for ANA and aPL suggested the possibility of SLE. The comments below will highlight the differential diagnosis of SLE relative to AAE, and the laboratory evaluation of low C4 levels in such clinical situations.
Low levels of multiple complement components indicate activation of the cascade (Figure 1). In contrast, an inherited deficiency results in a reduced level of one component. Because a C1q, C4, or C2 deficient patient often presents with SLE, distinguishing between genetic deficiency of C4 and classical pathway activation by immune complexes can be difficult. Further, multiple low levels can also be observed in C1-INH deficiency. Distinguishing among these possibilities has important diagnostic, therapeutic, and prognostic implications.
Both of our patients had unmeasurable C4 antigen by standard tests. This finding suggests either an inherited deficiency or marked classical pathway activation. The latter is most commonly due to immune complexes, as in SLE or mixed cryoglobulinemia. It may also be observed with antibodies to C1q (as in SLE or the hypocomplementemic urticarial vasculitis syndrome [HUVS]) (10–12) and with a deficiency of the C1-INH (hereditary angioedema [HAE] or AAE) (1, 13). The low C4 and C2 levels in patient 1 indicates accelerated turnover of the early components of the classical pathway. In SLE, such a marked reduction in C4 would lead to a low C3 as well. Therefore, the low C4 and C2, but normal C3 suggests fluid phase complement activation. In this situation, activated C1 cleaves C4 and C2, but the C3 convertase is inefficiently formed (13) and/or rapidly decayed (14). This profile is typical of HAE and AAE, but may also be observed in mixed cryoglobulinemia (14–18).
With the above considerations in mind, C1-INH level was determined. The marked reduction in C1-INH antigen and function established the diagnosis of AAE. The diagnosis of AAE then led to a search for autoantibodies to the C1-INH and an evaluation for malignancy.
AAE is usually first suspected on the basis of angioedema. In both of our patients, the lack of obvious angioedema confounded the differential diagnosis. In patient 1, the abdominal pain syndrome may have represented angioedema of the bowel. Patient 2 may have had 2 episodes of facial angioedema early in the disease course. Nevertheless, both patients have now been followed for over 2 years without angioedema developing. Several patients with a deficiency of C1-INH but without angioedema have been reported (1, 6, 17–25). There is also one report of AAE occurring in the absence of angioedema and in association with SLE, APLS, and an IgM monoclonal paraprotein (6).
AAE is commonly associated with low-grade B-cell lymphoproliferative disorders (1, 2, 5, 23, 26). Angioedema usually precedes recognition of the B-cell disorder. In this population idiotype–antiidiotype immune complexes were proposed to fix C1q and initiate the classical pathway (2). Classical pathway components and C1-INH would then be consumed by the sustained activation of the cascade (2, 27). This pathophysiologic explanation needs to be reconsidered because the autoantibodies produced inhibit the function of C1-INH (3, 8, 9). In one study of 13 patients with AAE, an autoantibody to C1-INH was eventually identified in 12 (3). Seven had a monoclonal band detected within 4 years of the diagnosis of AAE and in 6 the antibody to the C1-INH was of the same class and isotype as the paraprotein. Additionally, 5 of these antibodies bound to the C1-INH. Such a monoclonal antibody would prevent the stable association of C1-INH with C1r and C1s (28). Uninhibited C1s cleaves the C1-INH to a lower molecular weight (96 kd) nonfunctional form (7, 29) (Figure 2).
In patient 2, AAE coexists with the aPL and syndrome X. The latter condition represents a form of angina likely due to an impaired vasodilatation of the coronary arteries (30). The impaired vasodilatation and the absence of angioedema in the presence of C1-INH deficiency may share a common mechanism. Bradykinin plays an important role in coronary vasodilatation (31) and is one of the proposed mediators of the angioedema in C1-INH deficiency (32). Patient 2 could have defective generation or a reduced vascular response to bradykinin that caused syndrome X but prevented angioedema. Antibodies to phospholipids could interfere with kininogen by binding to the complex of kininogen-phosphatidylethanolamine (33).
In summary, the importance of these 2 case reports for the rheumatologist is several-fold. Angioedema need not be either a prominent symptom or even present in AAE. Because this symptom first suggests the diagnosis, AAE may be under recognized or confused with SLE. In cases of suspected C4 deficiency, laboratory evaluation should include a C3, C2, CH50, C1-INH and, in some cases, C1q, and antibodies to C1q and the C1-INH. An activation process can usually be ascertained based on the presence of multiple low components. A profile of low C4 and C2, but normal C3, should trigger the consideration of C1-INH deficiency. A low C1-INH then establishes the diagnosis of AAE. Making this differentiation impacts on treatment and followup, especially relative to lymphoproliferative surveillance. Finally, the presence of aPL may modify the expression of the angioedema as well as add to the diagnostic difficulties.